horizontal lines
Gigasheet Primary logo
  • crystalaryeh

Pivot Tables are Dead. Long Live Groups.

Groups are tremendously powerful. As you start to analyze the dataset you're uploaded, many of the questions you'll ask can probably be answered by Grouping; create a free Gigasheet account here to try.

In this blog post, we will show you how Groups work in Gigasheet—in this instance, we're working with vehicle sales data.

View the sample data and click View Only > Save A Copy to save it to your own account.

This video shows the process in action:

This synthetic data shows are nearly 4.7 million car sales based on craigslist. You see, each row represents one sales record, with the craigslist Region, the car manufacturer, the model, the car's year, the odometer, and on and on. The first thing you probably want to do is analyze this data by attribute, which will be done with Groups.

So, we can start with the Region. How many regions are there, and which has the most sales? That's answered simply by clicking on Region and then Group.

Now we see we have data from 404 separate regions. Columbus has the most sales, at 39688. Jacksonville is second at 39182, Spokane is third, Eugene, Orlando, etc. Clicking the arrow here lets you get to all the data.

To further understand this data, we can have Groups nested inside of other Groups. So what make of car gets sold the most per location? That's another Group, this time by Manufacturer. Now we see that in Columbus, Ford is the most followed by Chevy, then 2783 of Toyota, and so on. Collapsing Columbus, we can go to Orlando to see the results there. And again, if we wanted to further Group, we can split the ford sales by vehicle year. So, we see 2017 Ford's sold the most in Orlando. The most were 451 from 2018, then the 341 from 2019, and so on.

And again, easy to scroll down and now see, say, Nashville Toyota data.

Now say instead of answering what makes are sold by Region, we want to know what regions are sold by Manufacturer? That's as easy as dragging the Grouping breadcrumbs. Now we see 43 Groups representing the 43 values in the Manufacturer column. And we see Ford sold the most at 780835, and drilling in; we see a tie between 2016 and 2018 Hawaii Toyotas, both selling 429 cars. And of course, the year Grouping is there too.

Finally, we can learn more about this data via aggregation. So we can come to the bottom and choose for Odometer Median, and for the price, let's do Average. So now we're seeing on a Group by subgroup basis the Median odometer and average sale price for each of these records. We can see below in Jacksonville 2017 row that the average price was $13,000.

So hopefully, you will start to understand the power of Groups in Gigasheet. We think this ability to slice and dice rows and columns of a massive dataset is pretty cool. And we hope the familiarity of spreadsheet-like functions and workflows allows you to get to your answers quickly, no matter the size of your data. Gigasheet can handle files up to one billion cells of data. Sign up at gigasheet.com, free.

Recent Posts

Explore sample data from IMDB